Internet-audio streaming is evolving disparately from the
traditional radio broadcast model as computer-domain opportunities
increase. Allowing the end user to interact with the content of an
online experience is one such opportunity, and this interaction is
definitely making its mark in webcasting.
Browsing websites has always been interactive to some degree.
Following links from page to page requires user input. However, it
isn't until the user can send more information that content-based
interaction becomes possible. A good example of this would be with a
Figure 1. Customizable webcast among various streams.
Online services such as banking, shopping or gaming have even more
interaction. These services require the ability to uniquely identify
each remote user in a way that lasts between online sessions, unlike
the generic search engine example. Preferences that influence how the
service operates can now be set once and remembered indefinitely. From
the user's perspective, the Internet experience becomes customized at
the cost of anonymity.
Looking specifically at streaming audio services, Figure 1 shows one
possible arrangement. The player application or Web browser becomes
associated with aspecific user at the start of a session with a unique
user ID. From there, the player is automatically configured and
connected to the appropriate audio server. The user listens to one of
any number of preselected favorite streams, while the webcasting system
uses the information and user feedback paths to provide additional
non-audio content to, or acquire feedback from, the listener. It is
important to note that the blocks of the webcasting system represent
functionality, and that real-world implementations may use any number
of computers or other dedicated hardware to create the full system.
Spinner and Live365 use this approach.
The webcasting system in Figure 2 uses a slightly different
arrangement. The listener has a feedback path that extends all the way
back to the scheduler logic, which is able to generate and update a
unique playlist for that listener. Again, it is important here that the
webcasting system has a user database because each listener's profile
must also include scheduler-related preferences in addition to
operational preferences. Unlike Figure 1, the audio server here will be
generating a unique webcast for each listener in an almost on-demand
capacity. One company implementing this method is RCS.
Figure 2. Customizable webcast within a single stream.
In either system it is important to acquire the proper licensing
permissions and pay the appropriate royalty fees to the copyright
owners for any program material that is webcast. Currently, a statutory
license exists under the Digital Millennium Copyright Act (DMCA) to
cover webcasting systems, although it provides a number of constraints
that a listener-customizable service could easily violate. The
constraint of most interest here is that once the content of an audio
stream becomes interactive (as in Figure 2) the DMCA license would no
longer apply. It would be up to the webcasting service to acquire
individual license agreements from each copyright holder (the record
Nosé is president of NeoSonic Industries, Cleveland.